Abstract: We investigate musical genre classification for three forms of Marathi vocal music each with a distinct socio-cultural context, namely devotional, poetic and folk dance. We present a dataset of songs covering the three genres, and discuss their musical and acoustic characteristics. We consider timbre and chroma based features for the 3-way classification. We specifically examine whether the source-separated vocal and instrumental accompaniment components can help improve genre recognition accuracy over that obtained with acoustic features extracted from the original mix audio track.